A few days ago I was lucky enough to be able to do an on-phone interview with Scott Farrar the special effects supervisor and second unit director on Transformers: Revenge of The Fallen. You may not recognized Scott by name, but you have surely seen his work. Scott’s masterful resumé includes Star Wars VI- Return of The Jedi, Star Trek: The Wrath of Khan, Who Framed Roger Rabbit, Jurassic Park, Men in Black, Artificial Intelligence: AI, and many other classic films. His work on both the Transformer films are ground breaking achievements in special effects.
How much effects time did you have?
Scott Farrar: Well, on the first film — actually, each film took about a year and a half. We start with the script, and then we start breaking it down to figure out how many shots and how much work is going to be involved. After reading the script we realize, “Oh my gosh, there are forty to six new characters!”, and they have major speaking roles and they have to act a lot better then they did in the first one. Right away you see some pretty daunting challenges.
How do you make sure all the robots are different and unique in their own way? So when they are battling the audience can tell the difference, I imagine that’s tricky.
Scott Farrar: We know that in the first movie some of the characters were a little bit confusing, because some of their colors looked similar. That was a big note from fans and audiences. We tried very hard to make sure we had colors and shapes that were a little easier to tell apart. We try to make sure the fallen standout, and hopefully, the voice actors help the audience recognize some of the characters. It’s pretty easy for us to keep track of the characters, because each one has a name so right away we start calling them by their name when watching dailies. We can recognize that buffoon Starscream kissing up to Megatron! (Laughs)
Check out the rest of the interview below > ” />
How long generally does it take to model each robot, and which one took the most time?
Scott Farrar: Devastator — without a doubt — took the longest to model. He is goliath in proportion; he is one hundred and fifty feet tall compared to poor little Optimus who is only twenty five feet tall. Megatron is also around thirty-two feet. He’s composed of a thousand parts and he has like eight different gigantic road grating dirt movers, so he is huge and complex. For a simple model character, it would take about twelve weeks to model. That means just making all the shapes and putting them together to spin around to look at it. He doesn’t move yet at that point, and he has some rough paint on him or he may just be plastic grey shapes. Then the rigor goes in for about fifteen weeks rigging all the pieces together showing it as a skeleton frame so it can move around on the computer without all the pieces falling off. Then your kind of ready for shop production.
You start with the paint and textures, then you assign materials. We have a whole materials list that’s been developed using shades and software technology to assign each coding color or texture. Then your talking about twenty plus weeks until you can put a character into production and for Devastator we divided up in teams to build his various parts so then he could be combined together as one. He’s just painful from the stand point of rendering. It would take like forty-five minutes just to load him up on a screen just to look at him for one snap shot frame. Then it took seventy-two hours to a render a frame with him. (Sigh) It was on a new level of difficulty. Another interesting thing is that we used twenty terabytes disc space for memory for the first film and here its one-hundred and forty terabytes…
Now, is that number the final composition?
Scott Farrar: Yeah, that number is everything with all things considered.
What’s the big difference between rendering an IMAX shot?
Scott Farrar: Yes, probably the easiest way to describe it is to talk about an anamorphic film frame. While a film frame is just a thirty-five millimeter frame just like in a still camera, it’s the same size. An IMAX frame is on seventy-millimeter film, so it’s eight times larger. If you want to have same fine grain, which you do, you got to have eight times the information and space to render that shot. So that gives you an idea and then suddenly everything is at a much higher level to produce the IMAX sections of the film.
How long did it take to complete Devastator?
Scott Farrar: Oh gosh (laughs), Devastator was kind of in process for about eight months. It is not just that because it’s a complicated question. You know we build one of the dirt movers and then we work out what the transformation part might be and also how the pieces get plugged together. You find out something doesn’t work, so you have to go back and redo something. The camera changes and then you got to change the way the animation originally was.
Don’t you ever miss model making and crafting miniatures?
Scott Farrar: We actually still do that quite often, and we shot a number of miniatures for this film. When the aircraft carrier gets hit, we shot a lot of first scale flame elements. Also, when Wheel-Bot is crushing through the freeway, we built a miniature freeway and broke that apart. The Egyptian pillars that were Karnack’s temple we had to break apart, so we also used miniatures there. Also, we used them for lots and lots of dirt hits for the forest fight. We still do as much as we can, but we’re at a strange point in cinematic terms, where the simulations are getting much better. When you go to a huge scale like the render of breaking the pyramid apart, you’ve got to make a simulation of that, because there is no other way of doing it. You can’t do that in miniatures so that’s the cross over.
When you were working on Return of The Jedi and The Wrath of Khan did you ever imagine special effects coming this far?
Scott Farrar: Um, no (laughs). At the time, you always would think what you were doing was so ingenious and inventive. What’s interesting is that I’ve worked with people on this, and I always say its like were students how were always pushing forward to try to make things look better. It’s a collective thing that happens and sometimes you take big step forwards with ideas while other times it’s just little baby steps. Like Deep Impact, that was one of the first really big water shows that was done before The Perfect Storm. So it was like, “Oh my gosh, that was the hardest thing I have ever had to do in my life.” We were trying all these things that were huge movie concepts that were so difficult. We made some end roads, but not as much as the next guy. It was as if we were pioneers then.
How are you going to top all this with the third film?
Scott Farrar: Oh you just wait!
Maybe do ninety-two robots next time?
Scott Farrar: (laughs) Yeah, they are going to be on a football team next time! Well, it’s all in the end going to be script dependent just like every other film. The mythology of all these characters — the Transformers, the Autobots, and the Decepticons — is very deep if you look at the original source material. It’s just like Greek mythology with all these characters and this whole world. You could just go on and on with this, so I think that it’s really up to the writers and what they want to do with these mythic ideas. It’s really always about good versus evil. The characters in this movie have grown up a little bit, you know? Shia is growing up and he goes off to college where he is no longer just a boy with his robot car, its bigger then that. There are so many ways to go next time and there are a lot of great ideas that we haven’t touched yet.
Was it more important this time to give more detail to the robots facial expressions, so you they could convey more emotions?
Scott Farrar: Yeah, absolutely. When you see the film, you’ll see better examples of this. There is a character named Jetfire who is a pivotal character in the movie who moves the story along, and he’s involved with the actors. His facial performance and the emotions of his face needed to go up many notches of what he did with the first film. That’s true for several other characters including the ultimate bad guy who needs to look evil and menacing. It’s like Shakespeare drama, where you have real bad guys and you might be heading towards tragedy so you have to display emotion and anxiety. It’s pretty interesting to have this opportunity to be able to create this stuff and it’s a lot of fun.
How involved are you when it comes to the designs?
Scott Farrar: A lot of the original art comes from the production department in LA that gets up and running early on. We know all those people and we all work hand-in-hand. We’re heavily involved in that stuff. They do 2D-art with the front views and the side views of a robot, then we turn it into a three-dimensional shape. Most of the time, changes have to be made since the artist cannot possibly draw or paint in 3D. Once it goes to 3D, a lot of changes occur, and then I’ll have to run that stuff by Michael. After that, I tell Michael that we may have to change this or that since the robot will look ugly in front of a wide-angle lens. We’ll just thin the robot’s face, so it’s a lot of back forth when it comes to this stuff. That’s what makes this collaboration fun.
Transformers: Revenge of the Fallen hits theaters tonight (June 23rd) at midnight.